Gentleman Jon —Firstly an apology and correction. Due to Mike’s habit of off class mercing for random teams at Lan and playing Epsilon games using different steam accounts he wasn’t given the full credit for his damage in part 1. Taking this into account I can now correct part 1’s DPM figures, Mike does in fact have the highest damage output of any soldier at i49 regardless of heals (although still behind the top Demos). I can only beg forgiveness for portraying Mike as anything other than supreme.

War’s figures don’t get any correction, I checked those enough the first time round.

Whereas part 1 focused on damage and healing, this article will focus on kills and related stats.

Kills + assists / deaths

There’s loads of different ways of showing kill stats but to keep the number of graphs to an acceptable level I’ve gone for KAD, which also has the nice aspect of giving almost all players a positive number to keep them warm at night.

i49 KAD

Epsilon’s top two complement each other perfectly,  Numlocked doing the damage (and getting masses of kills) and  Stefan doing the cleanup with KnOxXx’s heal beam firmly up their (and Mike’s) backsides to net him the assists.  Bash and  Gear were obviously given a more suicidal role and Mike lands somewhere in between.

Team Immunity’s players take 5 of the top 11 places, their style clearly emphasising patient controlled encounters where they can establish a numerical advantage, guarding every frag carefully. They put team play and coordination ahead of overall aggression, using Aporia as the tactical sacrificial lamb to create openings, Yuki’s as a battering ram, or Sheep’s deadly sniper rifle.

B4nny’s high rating here in contrast to his mediocre DPM reflects a style based on having to take care of preserving his life over raw aggression (apart from some spectacular contrary examples), which as we’ve seen could be a consequence of relatively low heals. Overall it appears HRG aimed for the same kind of control that Immunity did, but without as much success. It’s hard to find positives for  Taggerung’s tournament.

The two Demo carries,  Brego and War, have mediocre KAD. The reasons for this are complicated, War has poor assist levels probably because no one else was cleaning up his damage and it’s hard to stay alive when the rest of your team are dead, as well as being forced to deathmatch by players cleaning you up. Also medic deaths on a team force aggressive suicidal choices. In a mirror to Team Immunity elem.v players take up 5 of the bottom 10 places. War and Brego’s high DPM but mediocre KAD numbers indicate aggressive play, but is it a personal style or the best they can do in the circumstances? Is the apotheosis of War at an end?

It’s a mixed picture for TCM, players who were unremarkable on DPM, Ryb and  Bybben, do better here, staying away from the spawn queue and providing the core control for the team.  Zebbosai by contrast had a tendency to give up his life to put out lots of damage, leaving it for others to clean up. It’s also worth remembering they played Epsilon for 3 extra maps which can’t have helped.

Infused have mediocre KAD as well as mediocre DPM, possibly a result of an attempt to balance play styles between aggression and control and adjust appropriately at the right moment. It’s unclear whether it was raw DM or timing/coordination that let them down.

Flippy does surprisingly well here, but the soldiers for the Frenchies in comparison to their excellent DPM numbers do poorly revealing their extreme aggro tactics from which Flippy benefits by assisting or cleaning up. Even Exon seems to be all out attacking although his DPM numbers weren’t spectacular.

Kills Per-Minute

i49 kpm

Back to the aggression stats and we can see that one of the keys to Epsilon’s success is that Numlocked combines durability and aggression filling first spot for KAD and second for KPM. Mike is also right up there, in part 1 we saw he received pocket level heals but his lesser KAD is indicative of his hybrid semi-roamer role. They don’t give each other many assists (part 3 article?) so they seem to be damaging different targets although how many assists KnOxXx sucks up and whether it obscures a relationship between Numlocked and Mike’s damage and kills isn’t clear. Assuming they are separate it’s easy to see why Epsilon are so strong.

War returns to top place, proving that he not only does the damage, he makes it count with kills.

Team Immunity’s results show a heavy dependence on two players to frag consistently fast and make things happen, Yuki being the battering ram or defensive perimeter, and Sheep cleaning up or getting headshots while the others wait for their chances to coordinate as a team and secure kills advantage. As noted later kill advantage counts for much more than damage.

HRG also have a methodical style but only Lansky is near the same class as the fastest fraggers and isn’t in the elite group.

Amongst the medics KnOxXx is the most successful with his ubersaw, and Ace manages a creditable performance, no doubt down to the number of attackers he found bearing down on him after tearing through his paper thin protection.

TCM’s lack of top performers seems to be a matter of being more willing to die and trust others to clean up, and they didn’t have an outstanding individual bridging both KPM and KAD categories like Numlocked. In fact it looks like only Yuki can match the status that Numlocked achieved of maximum aggression and effectiveness, but with about 50% more heals. Zebbosia was a top DPM soldier but in KPM he falls behind the elite group showing more willingness to go beyond the point of no return just to put out damage.

Who killed who?

There isn’t really space to show this graph so click to open it up in its full glory.

i49 who killed who

None of the numbers are adjusted for maps played or anything like that, it’s simply raw kills so teams meeting more than once and playing long matches will obviously have more kills against each other, hence Mike nearly racking up 50 kills against Jukebox. Anything over 10 is highlighted.

Medic Addendum – Uber Drops Per Game

i49 Drops Per Map

As this shows within a certain range taking risks with an uber doesn’t stand in the way of success. The uber drop rate didn’t get in the way of anybody doing well, but then again none of them were dropping as much as a single uber per map. Cross referenced with the Ubers per map stats from part 1 it appears  Fenrir and Ace weren’t getting time to build many ubers at all let alone drop them, so contrary to expectations very low drop rates can be a sign of poor performance.

Demomen – the ultimate test of skill

With Numlocked setting huge KAD numbers as well as fragging almost as fast as War it might be tempting to rate him as the top Demo in the tournament. However as everyone knows the ultimate test of demo skill is killing scouts with direct pipes. This chart shows the number of scouts killed with direct pipe hits per map.

i49 Demo pipes scouts

So I think that conclusively settles any debate.

Headshot Kills Per-Map

i49 headshots

As Sniper is a situational choice these stats aren’t really comparative and merely illustrate the quantity of sniping a team/player had success with. Sheep and Jukebox’s exploits are well known, but Flippy also inflicted plenty of head-exploding damage.

Amongst the top snipers Flippy showed the strongest preference for Medics while Sheep targeted more Demos. In contrast to the stereotypical sniper pick scenario scouts and soldiers constituted the majority of the most prolific snipers stats. Sheep and Flippy were most successful in the sniper vs sniper battles.

Tournament Backstabs

i49 backstabs

Successful spy plays were extremely rare, only Jukebox had semi regular success with it. In contrast to sniping the quantities are obviously far less but spies were able to pick one of the more valuable classes, demo or medic, with their first attack in higher proportion to total damage than snipers.

Damage vs Frags – Winning against the tide

From the logged round scores 11.5% of rounds were won by the team that did less damage than their opponents, but only 4.2% were won by the team with inferior kills. It’s far harder to overcome a frag disadvantage than a damage disadvantage.


Uneasy lies the head that wears a crown, War may feel his new title slipping although it’s fair to say you can’t get a good KAD without at least half decent team play. Still at least equal with b4nny, worst case scenario.

Epsilon owned 3 of the top 5 KAD spots and 2 of the top 5 KPM spots. Along with the most prolific healing it’s hard to see how any other team could possibly have competed. Given that, TCM get even more credit for making the final as close as it was.

Many of these metrics are only a nipple hair’s width between the top and mid table which shows how small the margins between winning and losing can be at the top.

Europe has too many demomen of a high standard.

Either the NA scene has gone off the boil or playing too much with the Quick Fix neuters your play style. Revenge NA super team for i52? Will the Euros have entered another malaise in a year’s time?